Visual simulation of a characteristic effect of an irregular phenomenon, such as smoke, fire, cloud, fog, or a waterfall, is an extremely challenging research subject in computer graphics. A specific shape and a characteristic effect are very difficult to be really described by using a conventional modeling method.
Nowadays, a particle system is usually used to present the characteristic effect of the irregular phenomenon, and the particle system can simulate a complex movement system. Position updating, death detection, and the like of particles in the particle system need to be processed by a central processing unit (CPU), and a graphics processing unit (GPU) displays the particles in the particle system according to the processing result. The process can consume a large amount of CPU time. Meanwhile, in the processing process of the CPU, the GPU needs to be in a lock waiting state. After the CPU generates particles and updates positions, the GPU can display the particles according to data updated by the CPU. Consequently, the processing efficiency is low.